Improved Feature Selection in Lorentzian Space

Yerzhan Kerimbekov, Hasan Şakir Bılge

2017

Closed access Metadata (CSL-JSON)
DOI 10.1145/3055635.3056582
Data confidence 70 medium
How this score is calculated

It is built from checkable properties of the record itself: whether the article has a DOI, how many independent sources confirmed its data, whether the required fields are filled in, and how recent the latest confirmation is. Disagreement between sources and a missing identifier lower it. A value of 80 or above means the record was published without further checking; 60 to 79 means it is published at reduced confidence, with a manual check advisable. A record verified by an editor by hand is never pushed below 90 by a recalculation. This scores our data about the article, not the article itself.

Abstract

Machine Learning is widely used in many areas because of its effectiveness in real-world problems. To make it more efficient a pre-processing is often applied. In this study, we suggest a preprocessing step for classification problem that is based on special Lorentzian space metric. This step consists of the novel Feature Selection in Lorentzian Space (FSLS) method and basic mathematical operations such as compression and rotation. The proposed FSLS algorithm uses the separability criterion which is redesigned according to the Lorentzian metric. The basic mathematical operations are used in order to make original data suitable for the Lorentzian space. Also, we suggest a new Lorentzian Classifier for Multiple Feature (LCMF).